LOSS of CONTROL OVER WATER **EXPLAINED**

[Zbip57]

Applying full throttle for a descent isn't needed in order to start the landing [...] a certain threshold which seems to be in the ballpark of a throttle value of -8300 to -8500.

Well that explains that.

Now I'm just wondering how I managed to get down to within 2ft of anything under the Mini, when it was still 25ft higher than my bench?

The circle marks where I believe the drone was at the time that auto-land was triggered. Maybe I was further back than I thought and got within 2ft of that treetop? But if that's the case, how did it manage to not land into the tree when the Mini then descended 15ft in auto-land mode?



And I really don't think the Mini was anywhere near that tree. Even if it was, the drone was at least 20ft higher than that treetop at the time.


Did you watch the screen recording video in post #58? Any clues in there?

It makes sense that auto-land was triggered when the VPS height showed only 2ft with me holding down the throttle. But I'm still baffled as to why the log showed a VPS height of only 2ft at that point.

 

[sar104]

Well that explains that.

Now I'm just wondering how I managed to get down to within 2ft of anything under the Mini, when it was still 25ft higher than my bench?

The circle marks where I believe the drone was at the time that auto-land was triggered. Maybe I was further back than I thought and got within 2ft of that treetop? But if that's the case, how did it manage to not land into the tree when the Mini then descended 15ft in auto-land mode?

View attachment 136503

And I really don't think the Mini was anywhere near that tree. Even if it was, the drone was at least 20ft higher than that treetop at the time.
View attachment 136505

Did you watch the screen recording video in post #58? Any clues in there?

It makes sense that auto-land was triggered when the VPS height showed only 2ft with me holding down the throttle. But I'm still baffled as to why the log showed a VPS height of only 2ft at that point.

The tree is the most likely candidate, unless there was a lot of spray in the air just below the drone.

 

[slup]

...Did you watch the screen recording video in post #58? Any clues in there?

Think we see the reason there in the clip ... alot of water smoke/mist coming from that waterfall, could have easily ended up on the sensor & tricked it.

 

[Zbip57]

The tree is the most likely candidate, unless there was a lot of spray in the air just below the drone.

Think we see the reason there in the clip ... alot of water smoke/mist coming from that waterfall, could have easily ended up on the sensor & tricked it.

I'm really hoping I wasn't stupid enough to blindly back the drone 2ft away from hitting a tree, but I can't 100% guarantee that wasn't the case.

So I'm going to stick with the spray/mist theory.

I guess that means, when flying in heavy mist, one should not hold the throttle down for more than 3 seconds at a time, and not close to full down.

Here's a video from a different angle showing how much spray/mist was coming off the waterfall.

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

 

[eEridani]

If it's any consolation, sometimes the vps indicator just messes up: note the jump from 30+ feet to 4 as my drone is ascending, T=~210.

 

[Zbip57]

Applying full throttle for a descent isn't needed in order to start the landing ... a couple preconditions needs to be met first though.

1. The VPS height should be 0,5m or lower
2. The throttle should in total be applied for descent longer than 3sec.
3. The throttle should sometime during the 3sec be positioned over a certain threshold which seems to be in the ballpark of a throttle value of -8300 to -8500.

I just confirmed this with my Mini.

Man, I was convinced it needed full down throttle to initiate landing. Not so.

Here, with the Mini hovering just off the floor and the VPS system preventing it going any lower, I tried moving the throttle down very slowly, and three times in a row it triggered Auto Landing well before reaching the throttle bottom stop.

Full scale throttle goes from +10,000 down to -10,000. The second test triggered landing at -8,500. I stopped the first two landings just by increasing throttle, and let the 3rd land itself with the throttle released (0 deflection).



Thanks for all the help, all of you guys! This forum is so helpful and educational.

 

[Pixeldawg]

First,
We see an elevation value on our remote control indicating feet above takeoff point.
We need to understand how a drone knows it elevation above ground.
Simply put it is barometric pressure
What is barometric pressure?
To simplify the discussion the following is important to know.

• The measurement of atmospheric air pressure is done with a device called a barometer
• Sea level pressure is expressed in millibars.
• Standard sea level pressure is 1013.2 millibars, 29.92 inches of mercury, 760 millimeters of mercury, or about 14.7 pounds per square inch.
• Pressure increases about 1 inch per 1,000 feet
• For a drone this equates to 10 feet for every 0.01 change in atmospheric pressure change.

EXAMPLE:
When I turn on my drone the electronics records and SETS atmospheric pressure in memory. This establishes my 0 foot elevation height.
As the drone increases in elevation the atmospheric air pressure changes and a calculation is performed against the set point to report the elevation in feet above ground (take off point).
This works great and as long as the atmospheric pressure that I am flying in does not change. The elevation reported and calculated is accurate.
What happens if the weather is changing which it is always doing? If after locking in the atmospheric pressure for 0 feet the drone flies and the actual ground air pressure changes by .01 inches (100th of an inch).
The drone will be flying with an elevation error of 10 feet.
So if flying over water you may think you are at 15 feet but in reality you may be at 5 feet above water assuming a change in pressure of only 100th of an inch in air pressure change.

Now let’s talk about temperature.
As temperature changes during the day the rate of heating causes air movement as air near the ground is warmed it rises. This rise in temp and movement of air may also contribute to the change in atmospheric pressure as air movement is what causes atmospheric pressure to change.

Now let us talk about drones and downward ultrasonic flight sensors.
When a drone finds itself above a solid surface the sensors provide a reflection in signal and a trigger to land and then a calculation is performed to control rate of landing.
Over land this is a simple safe landing.
Over water ultrasonic waves from these sensors PASS through water and are not reflected. (This assumes a flat water surface below the sensors.)
If the water has movement then some reflection may occur. This fluctuation may or may not be recognized by the sensor.

ors.
Since the drone cannot sense a surface it simply flies into the water.

Finally a combination in the above variables should be considered when flying over water.

So, a question for the OP here... I am fairly new to this and assumed much of this was done through GPS. It seems to me that most of the situations your have discussed here are for LOW LEVEL flights and not if you are say, 100 feet above the water. Although there may be a bit if difference in the altitude of the craft, it isn't nearly as critical as the example you mentioned between 15 and 5 feet. Is this correct?